Научный журнал Моделирование, оптимизация и информационные технологииThe scientific journal Modeling, Optimization and Information Technology
cетевое издание
issn 2310-6018

THE SHORTCOMINGS IN THE USE OF ARTIFICIAL NEURAL NETWORKS FOR SOLVING PROBLEMS OF BINARY CLASSIFICATION

Murashkin N.G.   Kostrova V.N.  

UDC 681.3
DOI:

  • Abstract
  • List of references
  • About authors

The article aims at identifying problems of the use of artificial neural networks for solving problems of binary classification. For solving problems of binary classification it is expected to classify samples that are already available to a certain class. A leading approach to the study of this problem is the algorithm of Levenberg-Marquardt, which allows to optimize the parameters of nonlinear regression models. As optimization criterion is adopted to the root mean square error of the model on the training set. It is proposed to accelerate the computation to apply the method to the elastic distribution. The materials of the article are of practical value to professionals who use artificial neural networks for classification tasks.

1. Choporov O.N. Metody analiza znachimosti pokazateley pri klassifikatsionnom i prognosticheskom modelirovanii / O.N.Choporov, A.N.Chupeev, S.Yu.Bregeda // Vestnik Voronezhskogo

2. Preobrazhenskiy Yu.P. Otsenka effektivnosti primeneniya sistemy intellektual'noy podderzhki prinyatiya resheniy / Yu.P. Preobrazhenskiy // Vestnik Voronezhskogo instituta vysokikh tekhnologiy. 2009. No. 5. pp. 116-119.

3. Choporov O.N. Metodika preobrazovaniya kachestvennykh kharakteristik v chislennye otsenki pri obrabotke rezul'tatov mediko-sotsial'nogo issledovaniya / O.N.Choporov, A.I.Agarkov, L.A.Kutashova, E.Yu. Konovalova // Vestnik Voronezhskogo instituta vysokikh tekhnologiy. 2012. No. 9. pp. 96-98.

4. Kruglov V. V. Iskusstvennye neyronnye seti. Teoriya i praktika. - 2-e izd., stereotip / V. V. Kruglov, V. V. Borisov. - M. : Goryachaya liniya-Telekom, 2002. – p. 382 .

5. Bodyanskiy E. V. Iskusstvennye neyronnye seti: arkhitektury, obuchenie, primeneniya: monografiya / E. V. Bodyanskiy, O.G. Rudenko. - Khar'kov: Teletekh,2004.- p. 369.

6. . Gavin, H. P. The Levenberg-Marquardt method for nonlinear least squares curve-fitting problems : : [electronic resource] / H. P. Gavin // Department of Civil and Environmental Engineering Duke University. - Access mode: http://people.duke.edu/~hpgavin/lm.pdf (date of request: 20.05.2017).

7. Riedmiller M. A Direct Adaptive Method for Faster Backpropagation Learning: The RPROP Algorithm : [electronic resource] / M. Riedmiller, H. Braun // University of Karlsruhe. - Access mode: http://deeplearning.cs.cmu.edu/pdfs/Rprop.pdf (Data dostupa: 20.05.2017).

8. Choporov O. N. The technique of formation of information database for multilevel monitoring and classification-predictive modeling / O. N. Choporov, O. V. Zolotukhin, I. Manakin, Bolgov S. V. // Vestnik of Voronezh Institute of high technologies. - 2015. - No. 14. - S. 19-24.9.

9. Chernov A.V. the Development of a taxonomy of predictive models the development of purulent-septic complications in puerperants / A.V. Chernov, V. Y. Brigadirovka, O. N. Choporov, and V. I. Chernov // System analysis and management in biomedical systems. - 2012. - T. 11. - No. 1. - P. 261- 266.10.

10. Babkin A. P. the Development of procedures for evaluating the severity of diabetic retinopathy in diabetic patients based on discriminant analysis / A. P. Babkin, V. G. Medentsev, A. G., Eng, O. N. Choporov // Bulletin of Voronezh state technical University. - 2005. - Vol. 1. - No. 7. - P. 97-99.

11. Choporov O. Technique of information database formation for carrying out multilevel monitoring and classificatory-and-forecasting modeling /Choporov O., A. Kurotova, I. Manakin // Information Тechnology Аpplications. - 2015. - No. 1. - P. 111-123.

Murashkin Nikita Gennadievich

Voronezh Institute of High Technologies

Voronezh, Russian Federation

Kostrova Vera Nikolaevna
Doctor of Technical Sciences Professor
Email: vn-kostrova@vivt.ru

Voronezh State Technical University

Voronezh, Russian Federation

Keywords: binary tasks, artificial neural networks, algorithm of levenberg-marquardt, algorithm of gauss-newton, method of m. riedmiller and g. brown

For citation: Murashkin N.G. Kostrova V.N. THE SHORTCOMINGS IN THE USE OF ARTIFICIAL NEURAL NETWORKS FOR SOLVING PROBLEMS OF BINARY CLASSIFICATION. Modeling, Optimization and Information Technology. 2017;5(2). Available from: https://moit.vivt.ru/wp-content/uploads/2017/05/MurashkinKostrova_2_17_1.pdf DOI: (In Russ).

59

Full text in PDF